1. 1 Dr. Sandro Sandri (President of Italian Association of Radiation Protection, AIRP) Head, Radiation Protection Laboratory, IRP FUAC Frascati ENEA – Radiation Protection Istitute sandro.sandri@enea.it 12th International Symposium on Radiation Physics 07 to 12 October 2012 - Rio de Janeiro - RJ THE RADIATION FIELDS AROUND A PROTON THERAPY FACILITY: A COMPARISON OF MONTE CARLO SIMULATIONS

2. CONTENTS S. Sandri - 12th International Symposium on Radiation Physics - Rio de Janeiro - RJ 2 The TOP-IMPLART Scope of the analysis Simulation model The computer codes Results Discussion and conclusion

3. TOP-IMPLART accelerator TOP-IMPLART are the acronym of Terapia Oncologica con Protoni (Oncological Therapy with Protons) and Intensity Modulated Proton Linear Accelerator for Therapy The first 7 MeV module of the accelerator, is already installed and has been tested Additional modules will be added leading proton energy to 30, 70 and 150 MeV In the final layout the bunker will be 30 m long and 3 m wide 3 S. Sandri - 12th International Symposium on Radiation Physics - Rio de Janeiro - RJ

4. COMPUTER CODES The principal subject of the current work is the analysis of the performance of two different computer codes both based on the Monte Carlo algorithm: FLUKA (FLUktuierende KAskade) and MCNPX (Monte Carlo N-Particle eXtended) Info on the web sites: www.fluka.org mcnpx.lanl.gov S. Sandri - 12th International Symposium on Radiation Physics - Rio de Janeiro - RJ 4

5. SIMULATION MODEL The model has been developed to simulate a 150 MeV proton beam hitting a water phantom of cubic form, 32 cm thick (32x32x32 cm 3 ) with 2 mm plexiglass walls and located in front to the kapton membrane, 50 µm thick, that seals the vacuum chamber of the accelerator Between the kapton membrane and the phantom there is a 2 cm air gap The cross section of the proton beam reaching the kapton membrane has the maximum dimension of 7 mm (in x and y directions) S. Sandri - 12th International Symposium on Radiation Physics - Rio de Janeiro - RJ 5

6. SECONDARY PARTICLES SECONDARIES FLUKA RESULTS PROMPT RADIATIONRADIOACTIVE DECAYS 4-HELIUM1.3456E-01 (23.0%)9.9215E-04 ( 0.9%) 3-HELIUM6.2313E-03 (1.1%) TRITON3.2014E-03 (0.5%) DEUTERON1.1862E-02 ( 2.0%) PROTON2.4968E-01 (42.8%) ELECTRON 6.6010E-03 ( 5.8%) POSITRON 4.8514E-02 (42.8%) NEUTRIE 4.8514E-02 (42.8%) ANEUTRIE 6.5313E-03 ( 5.8%) PHOTON7.0562E-02 (12.1%)2.1726E-03 ( 1.9%) NEUTRON1.0773E-01 (18.5%) TOTAL5.8383E-01 (100.%)1.1332E-01 (100.%) S. Sandri - 12th International Symposium on Radiation Physics - Rio de Janeiro - RJ 6 Several secondaries are generated in the inelastic interactions of the beam protons with the target components (plexiglass and water). Both the codes were able to follow the different produced particles and provided different kind of related results. FLUKA for examples provided the following table per beam particle

7. NEUTRON PRODUCTION S. Sandri - 12th International Symposium on Radiation Physics - Rio de Janeiro - RJ 7 The comparison of the data for neutron production shows a reasonable agreement between the two codes. However using the libraries in MCNPX the neutron yield is about 7% higher MCNPX METHODMCNPX NEUTRONS PER PROTON Bertini model 1.0485E-01 La150h 150 MeV Los Alamos proton Library1.1192E-01 ENDF70prot 150 MeV ENDF proton Library 1.1239E-01 FLUKA NEUTRONS PER PROTON 1.0773E-01 2,6%

8. PARAMETERS OF COMPARISON the comparison of the code concentrated on the following fluence results: Proton fluence in the target and in air around the target Neutron fluence in the target and in air around the target Photon fluence in the target and in air around the target S. Sandri - 12th International Symposium on Radiation Physics - Rio de Janeiro - RJ 8

9. FLUKA proton fluence In FLUKA the spatial distribution of a quantity can be reported in a 2d chromatic picture MCNPX doesn’t have this capability S. Sandri - 12th International Symposium on Radiation Physics - Rio de Janeiro - RJ 9 Water phantom

10. MCNPX proton fluence MCNPX, Proton fluence in air, 100 cm after the target The total proton flux of about 1 10 -8 (8,44% uncertainty) is the same of FLUKA S. Sandri - 12th International Symposium on Radiation Physics - Rio de Janeiro - RJ 10

11. Photon fluence FLUKA MCNPX 11 S. Sandri - 12th International Symposium on Radiation Physics - Rio de Janeiro - RJ Discrepancies in the results for photons are mainly due to different units and scaling

12. Neutron fluence FLUKA MCNPX 12 S. Sandri - 12th International Symposium on Radiation Physics - Rio de Janeiro - RJ Due to the different units, the qualitative path only can be compared in the graphs, showing a good agreement

13. FLUKA, Neutron fluence, spatial distribution Neutrons are more intense in the forward direction, as foreseeable S. Sandri - 12th International Symposium on Radiation Physics - Rio de Janeiro - RJ 13

14. CONCLUSIONS Both computer codes used in the simulation are well suitable to be applied to the analysis of the secondary radiation produced by the proton beam of the TOP-IMPLART accelerator While MCNPX seems to be more flexible in the data library selection and update, FLUKA can provide a more complete output in term of graphical detail Another advantage of MCNPX is the availability of versions developed to run on the world wide diffused Windows™ personal computer, on the other hand FLUKA can be installed on a pc with Linux system The results obtained with the two codes showed a good agreement for the fluence vs energy spectra of the neutrons (the main secondary radiation) In conclusion both the codes are appropriate for the specific calculation and the selection should be mainly based on the hardware and operative system availability, and on the specific skilfulness of the users S. Sandri - 12th International Symposium on Radiation Physics - Rio de Janeiro - RJ 14